Robust, fragile & anti-fragile¶
Energy and information¶
Information and energy are intimately related for all physical systems because information has to be written on some physical substrate which always comes at some energy cost (Landauer 1961; Bennett 1982; Leff and Rex 1990; Berut et al. 2012; Parrondo et al. 2015).
Brains are physical devices that process information and simultaneously dissipate energy (Levy and Baxter 1996; Laughlin et al. 1998) in the form of heat (Karbowski 2009). This energetic cost is relatively high (Laughlin et al. 1998; Aiello and Wheeler 1995; Attwell and Laughlin 2001; Karbowski 2007), which is the likely cause for a sparse coding strategy in neural circuits (Balasubramanian et al. 2001; Niven and Laughlin 2008).
Experimental studies (Shulman et al. 2004; Logothetis 2008; Alle et al. 2009), as well as theoretical calculations based on data (Harris et al. 2012; Karbowski 2012), indicate that fast synaptic signaling, i.e. synaptic transmission, together with neuron’s action potentials are the major consumers of metabolic energy. This type of energy use is of electric origin, and is caused by flows of electric charge due to voltage and concentration gradients and subsequent pumping ions out to maintain these gradients (Attwell and Laughlin 2001; Karbowski 2009). This very pumping of electric charge requires large amounts of energy.
Karbowski, J. 2021. Energetics of stochastic BCM type synaptic plasticity and storing of accurate information. Journal of Computational Neuroscience 49: 71-106